In a monitoring and controlling system that monitors and controls, for example, a building facility or a plant facility, communication devices, which have various functions such as information collection functions and control functions, are connected as nodes via a communication network, and a central monitoring device separately monitors and controls each facility based on information sent from the nodes. Such a monitoring and controlling system uses Ethernet as the communication network.
When using Ethernet, multiple nodes are fundamentally connected in a star configuration, wherein each node is connected to a hub, a switch, or the like. Although a star configuration is suited to relatively small scale office environments, it is not necessarily suited to large scale facilities, such as building facilities or plant facilities. This is because, in a star configuration, each node must be connected to a hub or a switch via separate wiring, and, if the nodes are installed over a wide area, the wiring that connects these nodes becomes complicated, which in turn increases the workload involved with the wiring work and maintenance.
An Ethernet switch that connects each node via Ethernet using a ring configuration has been proposed. This ring Ethernet switch is a relay device that connects multiple nodes in a ring via bridge wiring using network control functions such as a spanning tree protocol (STP; Spanning Tree Protocol/IEEE 802.1D), which resolves communication errors generated by a ring topology that exist in a communication path, and rapid STP (RSTP; Rapid STP/IEEE 802.1w), which improves on STP. Providing a ring Ethernet switch for each node makes it possible to connect multiple nodes in a ring configuration and to achieve system redundancy.
FIG. 7 is an example of a configuration of a typical ring Ethernet system. Here, multiple nodes N are connected to a ring L via ring Ethernet switches. Normally, the network control function, such as RSTP or STP, installed in each Ethernet switch selects one root node R from among the nodes connected in a ring, network control information in the form of so-called bridge protocol data units (BPDUs) is exchanged between the root node R and the other nodes, and thereby an active communication path configured in a tree topology is established based on the node-to-node ring costs.
At this time, ports of nodes in unnecessary paths, which are those paths other than the active communication path, are blocked and established as backup communication paths in the event of a failure.
In the example shown in FIG. 7, the path from the root node R to a node N1 could be either counterclockwise or clockwise. If the cost of the counterclockwise direction is lower than that of the clockwise direction at this time, then the counterclockwise path will be selected as the active communication path. Accordingly, the path from the node N1 to a node N2 becomes an unnecessary path, and it is blocked at either one of its end points, namely the node N1 or the node N2. Consequently, the original ring L configured in a ring topology is modified to a tree topology that comprises two branch paths, one from the root node R to the node N1 and one from the root node R to the node N2.
Thereby, even in a network that is physically configured in a ring topology, the generation of data loops is avoided. In addition, if any of the nodes becomes unable to receive the BPDUs transmitted periodically from the root node R, then it can be determined that a failure has occurred in the path between the root node R and the relevant node. In such a case, a restructure request is transmitted from the relevant node to the root node R in the reverse direction. In response to receiving this restructure request, blocked nodes are unblocked. Thereby, a new communication path is constructed using the backup communication path that was being blocked.
Accordingly, in the example shown in FIG. 7, if a failure occurs at a point P, a restructure request is sent from a node N3, the node N1 unblocks a point B, and thereby a new path is constructed from the root node R to the node N3.
If a large scale Ethernet network with a single ring is implemented in, for example, a building facility or a plant facility using ring Ethernet switches, then, because all of the nodes share a single ring, the reliability of the system as a whole decreases.
As an example of technology to solve such problems in the conventional art, a method (e.g., refer to Japanese Unexamined Patent Application Publication No. 2006-174422 and the like) has been proposed wherein the nodes are divided into multiple subrings and connected, and the subrings are connected to one another using general-purpose switching hubs. Thereby, risk is distributed among the subrings, which improves system reliability.
1
In this conventional art, physically independent subrings are connected by general-purpose switching hubs, which makes it possible to communicate between the subrings not only user data but also network control data. Namely, one domain is formed not only for data communication but also for network control.
Accordingly, in the conventional art, when a failure that occurs in an arbitrary subring propagates to another subring, the network control function performs an operation that restructures the communication path even in the normal subring; consequently, data communication to and from the nodes connected to the normal subring is temporarily inhibited, which is a problem.
FIG. 8 is an example of the configuration of a ring Ethernet network according to the conventional art. Here, the network comprises three subrings L1-L3, each of which is connected to a switching hub. The switching hub has a function that corresponds to the network control function (e.g., RSTP or STP) installed in each node and transfers BPDUs and restructure requests used in the network control function among the subrings L1-L3.
Thereby, from the standpoint of the network control function, the three subrings L1-L3 are regarded as one ring, wherein BPDUs transmitted from the root node R of the subring L1 are also transferred to the other subrings L2, L3.
Here, if a failure occurs at, for example, the point P in the subring L1, then a restructure request is transmitted from the node N3 to the node N2, and the point B between the node N1 and the node N2 is unblocked. Thereby, a new communication path is constructed from the root node R to the node N1 by using the backup communication path, which was being blocked, between the node N1 and the node N2.
To construct a new communication path, a restructuring operation is performed at this time even in the nodes connected to the subrings L2, L3, which are subrings other than the subring L1 at which the failure occurred, and consequently data communication to and from the nodes connected to the normal subrings is temporarily impeded.
The present invention was conceived to solve these problems, and it is an object of the present invention to provide a ring connection control circuit, a ring Ethernet switch, a ring Ethernet system, and a ring connection controlling method that can maintain data communication with a normal subring even if a failure occurs in an arbitrary subring.